DEV Community: Md. Lutful Hasan

Gift of Tech: A Custom Build Over the Wheels

Md. Lutful Hasan — Fri, 05 Sep 2025 03:27:52 +0000

Working in PCB store is fun, see lots of customers, experienced their thought is quite amazing. In PCB Store, we basically offer custom and pre-built PCs. So, in the store, we witness many things related to builds. Last month, in the store, we got requested for a custom build. But here is story.

The context is the son is a typical teenager who loves bikes and wants to ride with his fellow homies. After his college graduation, the son wants a higher CC bike from his father as a gift. But the PC-enthusiastic father has a different plan in mind. Father gifts him a top-of-the-line gaming PC instead of a higher CC bike.

So, as it's a gift from a father to his son, we take time and build a PC with the high-end components from our store shelves.

Not only that, but we also published a video about this build in our PCB YouTube channel online, to appropriate this father who decided to not give the bike to this son.

Back to the build:

As their budget was slightly high, we decided to use the best and latest component from our store.

Firstly, choose AMD Ryzen 9 9950X3D as a processor, as his top priority is gaming. As Ryzen 9 9950X3D is part of the Zen 5 architecture, it will give perfectly balanced output in the case of gaming and productivity. So, even in his higher study, he can do his additional research and others' work without any problems.

In the GPU section, we give them our latest and high-end GPU card from our shelf, the ASUS ROG Strix GeForce RTX 5090 Platinum, the fastest gaming GPU right now. Not to say much about it in here, 'cause all know what kind of beast it is in the gaming section. But the most amazing part of this GPU card, which many reviewers skip and do not mention in their video, is their GPU Tweak 3 software. With this software, you can check the temperature of different parts of the GPU. Also, you can overclock the GPU from this software. Also can under-volt and adjust fan speed.

Now, we need a motherboard to fit all the things. As a motherboard customer, choose ROG Crosshair Hero X870. It's a good choice with Ryzen 9 9950X3D and 5090, but here is a catch. We set RAM in the A2 and B2 slots, as we have PCIe lane reductions and give half of its power to this motherboard.

Gaming Performance Test

As it's a gaming-focused build, we test with 2 games, Black Myth: Wukong and Cyberpunk 2077. We test these two games in two gaming configuration environments. In one environment, we focused on rasterisation in native resolutions. And for another test, we enable ray tracing and do a rasterisation test in the cinematic preset, which may be a little bit overclocked.

In Black Myth: Wukong, in 1440p, the card gives us a decent playable framerate, but in 4k, with ray tracing in native resolution, the framerate drops continuously; sometimes we get 30 FPS.
Here I am attach two image of our results

In Cyberpunk 2077, we get decent good framerate in both test in both resolution . In 2025, it's still difficult to use native resolution in Ray-tracing. Also, attaching two image and results of our testing

In here I just share a glimpse of our daily routine. We do lot a stuff and it just a tip of the ice-burg.

Building PC for "Private AI Infrastructure"

Md. Lutful Hasan — Sun, 31 Aug 2025 18:51:01 +0000

When it's comes to handling sensitive data and generating highly specific outputs, nowadays most of the company, lab or researchers, specially in like sectors medical, healthcare, finance, trying to develop their own custom made AI Agent or model, tailored to their unique requirements and build their AI infrastructures.

Inside PC Builders Bangladesh office, we often get offer or request to build others server, PC and list go on.

Last month, we got a request to build a Private AI PC (Infrastructure) to maintain their research work with their private data. So, the team buckle up and start this project. It looks simple but the collecting the Nvidia GPU is tough for logistics team. Total cost of this build $16,930 USD. Hence set-up cost a lot but in long-term, it gives lower operational cost to them.

Click here you can watch the entire build

So, what component use to build this PC and why?

we use AMD Ryzen Threadripper 7980X as processor. This is perfect for their job. 'Cause Threadripper 7980X offers top-tier multi-threaded and PCIe lane capacity, which is absolute for AI servers, especially those combining many GPUs. In budget, it offers performance per dollar compared to many Intel Xeon and AMD EPYC CPUs when it comes to AI and model training workloads.

In any AI Infrastructures, the most important thing is GPU, specifically NVIDIA GPU. NVIDIA GPUs considered essential assets for AI and machine learning tasks. In this build we use, NVIDIA RTX A6000, which is suitable option in my opinion, as it gives massive VRAM capacity, which very compatible to handle larger models (30B+ parameters with quantization), bigger datasets, or higher batch sizes, without memory limitations(what client needed). Also, ECC memory and workstation drivers offers Professional-Grade Reliability. want to use NVIDIA Blacwall GPU, but it still not available in consumer market. Maybe, primarily NVIDIA only offers the giant company, who pays well to them.

Then come to the Motherboard. In this section selected Asrock TRX50 WS, a good match for pairing with the Ryzen Threadripper and RTX A6000 GPU, as it offers up to 7 PCIe 4.0 x16 slots, critical for multi-GPU AI training setups.

In cooler section, try Abee Stem 360. I think it's an not bad choice,as it has special metal water cooling head known for its durability, also it good choice for maintaining the high thermal output of powerful components like the Threadripper 7980X.

In the entire build, the cheapest components is PC-Case. In my view as long as it offers excellent airflow to keep all hardware cooled effectively , it's OK.

At a glance of all components we used:
Processor: AMD Ryzen Threadripper 7980X
GPU: RTX A600
Motherboard: Asrock TRX50 WS
Ram: Samsung 32GB ECC Server Ram (great choice for this build)
Memory: WD BlackSN7100 2TB
PC Case: Corsair 3500X
Power Supply: Thermalright TG1350 1350W 80 Plus Gold Power Supply
Cooler: Abee Stem 360

An Ideal PC for "Private AI Infrastructure"(?)

Md. Lutful Hasan — Sun, 31 Aug 2025 03:47:19 +0000

In 2025, almost every week some new AI model resealing in Hugging Face. Most of which are fine-tuned versions of large language models such as ChatGPT, Llama, or other open-source variants.

However, when it's comes to handling sensitive data and generating highly specific outputs, nowadays most of the company, lab or researchers, specially in like sectors medical, healthcare, finance, trying to develop their own custom made AI Agent or model, tailored to their unique requirements and build their AI infrastructures.

Focusing on Privacy and Security is the main reason to adopt this practice, especially for companies in heavily regulated sectors like healthcare and finance. Also, building in house AI infrastructures helps in cost optimization. Initial setup might cost a lot but Private AI gives lower long-term operational costs, optimized infrastructure investment, and reduced dependency on external providers. It also offers higher accuracy, shameless integration, better data governance, which help to achieve more reliable, context-aware results that directly enhance business operations.

But when I search online there are relatively few videos and blogs. Most of them are, some tech reviewer review big giant corporation's pre-build AI laptop or desktop server what they offers to consumer. Very few actually build this type of PC from scratch. But the adopting the private AI Infrastructures culture increases day by day.

Fortunately, there are exceptions. In YouTube I found a channel named PC Builders Bangladesh, with video, where they build a PC specifically designed for private AI infrastructures targeting sectors like healthcare and finance They give a huge efforts to build this. Total cost of this build** $16,930–$17,400 USD**. Their content provides practical insights for organizations and individuals interested in building their own AI infrastructure with moderate budget.

Here you can watch the entire build-

So, what component they use to build this PC?

They use AMD Ryzen Threadripper 7980X as processor. This is the perfect things for the job they want to do. 'Cause Threadripper 7980X offers top-tier multi-threaded and PCIe lane capacity, which is absolute for AI servers, especially those combining many GPUs. It offers superior performance per dollar compared to many Intel Xeon and AMD EPYC CPUs when it comes to AI and model training workloads.

In any AI Infrastructures, the most important thing is GPU, specifically NVIDIA GPU. NVIDIA GPUs considered essential assets for AI and machine learning tasks. In this build they use, NVIDIA RTX A6000, which is suitable option in my opinion. It gives massive VRAM capacity, which very compatible to handle larger models (30B+ parameters with quantization), bigger datasets, or higher batch sizes, without memory limitations.With ECC memory and workstation drivers, it offers Professional-Grade Reliability. In video, they said, they want to use NVIDIA Blacwall GPU, but it still not available in consumer market. Maybe, primarily NVIDIA only offers the giant company, who pays well to them.

Then come to the Motherboard. In this section they selected Asrock TRX50 WS, a solid match for pairing with the Ryzen Threadripper and RTX A6000 GPU, as it offers up to 7 PCIe 4.0 x16 slots, critical for multi-GPU AI training setups.

Another cool part of this build is cooler. They try** Abee Stem 360**, an ideal choice for this type of build. Why ideal? As it features special metal water cooling head known for its durability and significantly improved heat dissipation efficiency, which help to maintain the high thermal output of powerful components like the Threadripper 7980X.

In the entire build, the cheapest components is PC-Case. But this is acceptable as long as it offers excellent airflow, which is critical to keeping all hardware cooled effectively and maintaining system stability during intensive AI workloads.

At a glance of all components they used:

_Processor: AMD Ryzen Threadripper 7980X

GPU: RTX A6000

Motherboard: Asrock TRX50 WS

Ram: Samsung 32GB ECC Server Ram (great choice for this build)

Memory: WD BlackSN7100 2TB

PC Case: Corsair 3500X

Power Supply: Thermalright TG1350 1350W 80 Plus Gold Power Supply

Cooler: Abee Stem 360
_

As PC enthusiastic and working with AI,ML, I think they tried their best and they also done some great build. I'll commanded this channel if you want to appreciate some awesome build.

How to run "C++" program in Zed Editor.(for noob)

Md. Lutful Hasan — Thu, 31 Jul 2025 17:43:01 +0000

If you're reading this article, you're likely already familiar with the Zed Editor — a Rust-based, open-source, multiplayer code editor .As an open-source enthusiast, I always try to use and support open-source tools, software or anything, and Zed stands out as a truly next-generation code editor. Designed for speed and collaboration, it delivers an ultra-fast, intelligent, and responsive coding experience by leveraging multiple CPU cores and GPU acceleration.

When I first tried to execute my C++ program in Zed, I encountered a few challenges. Unlike traditional IDEs, Zed is largely script-driven, meaning you need to configure custom tasks via settings to execute programs. Since there’s limited beginner-friendly documentation on this, I wanted to share my experience to help others navigate this process. In this article, I’ll Walk through how to run a C++ program in the Zed Editor, step by step. I am using Fedora 42 as an operating system while testing that script.

Setup C++ Compiler

First, ensure you have a C++ compiler installed, such as g++ or clang++ in your system

Configure Clangd in Zed (Optional for language support):

Zed uses clangd as the language server for C++ support and formatting. You can configure the path to your clangd compiler in settings.json: use this type script.
For accessing settings.jso, open settings from Zed.


"lsp": {
  "clangd": {
    "binary": {
      "path": "/path/to/clangd",
      "arguments": []
    }
  }
}

Create a Task to Compile and Run:

Use Zed's "tasks" feature to automate compiling and running your C++ code. For this follow the steps,

Open the Tasks Configuration:
1.In Zed, open the Command Palette by pressing Ctrl+Shift+P (Linux) or Cmd+Shift+P (Mac).

2.Type zed: open tasks and select it to open the tasks configuration file (tasks.json). This can be a global file (~/.config/zed/tasks.json) for all projects or a local file (.zed/tasks.json) for the current project.

Add a Task Entry:
Inside the JSON array, add a new object representing your task. For example, to compile and run a C++ program:

[
  {
    "label": "Compile, Run, and Extra Commands",
    "command": "g++ $ZED_FILE -o $ZED_STEM && ./$ZED_STEM && echo && echo 'Program finished' && echo",
    "use_new_terminal": false,
    "allow_concurrent_runs": false,
    "reveal": "always",
    "hide": "never",
    "shell": "system"
  }
]

Here, $ZED_FILE is the current C++ source file.
$ZED_STEM is that filename without extension, used as the executable name.
The commands run sequentially:
1.Compile the C++ file with g++.

2. Run the produced executable.
3. Print a message (echo 'Program finished').

4. List files in the directory (ls).

If want to extra command to add
Define a new task with a command string that includes multiple shell commands separated by && or ; depending on whether you want them to stop on failure or run regardless

Save the File:
The tasks.json file will save your task.

Run the Task:

Open the Command Palette again.
Type task: spawn and select your "Compile and Run C++" task.
The output of the compilation and running will appear in the integrated terminal.

Now,You can run C++ program in Zed text editor.

How to set Keymap for execute the program:

If you want to set key bindings for execute the program, then maybe these steps helpful for you.

Open your Zed keymap configuration:Open the command palette with Ctrl+Shift+P (or Cmd+Shift+P) and typeZed: Open Keymap.
Add a binding for your custom shortcut:
then edit keymap.json with this template. Here I write for the snippet the first one

[
  {
    "context": "Workspace",
    "bindings": {
      "shift-e": ["task::Spawn", { "task_name": "Compile, Run, and Extra Commands" }]
    }
  }
]

The task_name string must exactly match the "label" in your tasks.json (including capitalization, spaces, and punctuation).
"shift-e" is my desired shortcut.
Save the file and restart Zed.
Now, pressing shift+e will run your entire C++ program with a single keystroke, just like VS Code's "Run" button, without needing to use the Command Palette

How to create snippet

Follow these steps, to create any snippet for c++ program.

Open Zed’s command palette (Ctrl+Shift+P or Cmd+Shift+P).
Select snippets: configure snippets.
Choose C++ (c++.json) or create if it doesn't exist
Paste the above JSON snippet inside the file (make sure it’s valid JSON with commas if you add multiple snippets).
Save the file.
In C++ files, type cppmain and press tab to expand.

Let, I am here writing a simple c++ snippet for main function:

{
  "C++ main template": {
    "prefix": "cppmain",
    "body": [
      "#include <bits/stdc++.h>",
      "using namespace std;",
      "",
      "int main()",
      "{",
      "   $0 ",
      "    ",
      "    ",
      "    cout << endl;",
      "    return 0;",
      "}"
    ],
    "description": "C++ main function template with bits/stdc++.h"
  },
  "Another snippet": {
    "prefix": "as",
    "body": ["// Additional snippet here"],
    "description": "Example snippet"
  }
}

So, what we do here:

prefix: "cppmain" — This is the trigger text for the snippet.
body: An array of strings each representing a line in the snippet.
$0 is correctly placed to mark the final cursor position inside the main function.

Just make sure:

This snippet object is inside the main JSON snippet file for C++ (usually cpp.json).
If you have multiple snippets in the same file, remember to separate them by commas.